http://ijhpm.com Int J Health Policy Manag 2021, 10(4), 192–200 doi 10.34172/ijhpm.2020.44 Original Article Education and Experience as Determinants of Micro Health Insurance Enrolment ID Basri Savitha1* , Subrato Banerjee2,3 ID Abstract Background: India faces a formidable challenge of providing universal health coverage to its uninsured population in the informal sector of the economy Numerous micro health insurance (MHI) schemes have emerged as health financing mechanisms to reduce medical-illness-induced poverty Existing research shows that the purchase of health insurance is most likely to be determined by health status, expected healthcare expenditure, and past health experiences in addition to socio-economic variables We add to the understanding of various factors influencing enrolment in MHI from an Indian perspective Methods: A survey was carried out to collect quantitative data in three districts in the state of Karnataka, India Results: We show that education does not matter as significantly as experience does, in the determination of new insurance purchases In other words, the importance of new insurance is not understood by those who are merely educated, but by those who have either fallen ill, or have previously seen the hazards of usurious borrowing Conclusion: Our study provides deeper insights into the role of usurious borrowing and past illness in determining insurance purchases and highlights the formidable challenge of financial sustainability in the MHI market of India Keywords: Micro Health Insurance, India, Illness, Usurious Borrowing, Education Copyright: © 2021 The Author(s); Published by Kerman University of Medical Sciences This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited Citation: Savitha B, Banerjee S Education and experience as determinants of micro health insurance enrolment Int J Health Policy Manag 2021;10(4):192–200 doi:10.34172/ijhpm.2020.44 Article History: Received: April 2019 Accepted: 15 March 2020 ePublished: April 2020 *Correspondence to: Basri Savitha Email: savitha.bs@manipal.edu Key Messages Implications for policy makers • Adverse selection and the consequential financial non-sustainability must be curtailed through a scrutiny of the risk-profile of prospective clients • Scheme administrators could collect data on illness and borrowing habits that concern the social capital of rural communities • In addition to the compulsory enrolment of all family members, a waiting period of one month could be enforced • Instead of risk-rating on the part of the community, one could adopt the sliding scale methodology to determine premiums (and consequently, charge higher premiums for high-risk individuals) Implications for the public For impoverished households, income and education may not be obstacles to enrolment The experience of illness and its repercussions on the household (giving rise to ex-post regret for not being insured), however, has a significant influence on the decision to (eventually) enrol in micro health insurance (MHI) While insurance (of renewed policy) claims from the MHI scheme reduced out-of-pocket expenses (OOPEs), those newly insured had a comparably higher OOPE, necessitating higher borrowing from multiple sources including usurious and non-usurious credit Since usurious loan has severe consequences on the financial well-being of any household, the non-insured joined Sampoorna Suraksha Programme (SSP, which is formally explained later) to mitigate the impact of future (adverse) health shocks Background A considerable amount of emphasis is being put on the need to educate consumers on the merits of a product,1 “particularly in this age of rampant misinformation, a disinterested public, or the genuine possibility that customers simply don’t believe they need a given product.” The lack of a feeling of necessity (noted at the very end of the previous quote), often presents itself as a hindrance to insurance buying For example, in a recent study,2 it is observed that half the respondents indicated Full list of authors’ affiliations is available at the end of the article confusion about their health insurance plans, often leading to the delay or a complete foregoing of medical care Therefore, targeted outreach and education programmes for buyers of insurance products are often recommended to fill in these ‘knowledge deficits.’ Indeed, many recommendations have been offered by both academics and industry to improve financial education in general.3 The central motive of this paper, is thus, to re-examine the belief that those who are educated are more likely to buy insurance We emphasize that Savitha and Banerjee there are more important determinants of insurance buying in comparison with insurance literacy We first stress, through our data from households in Karnataka, that insurance buying behaviour does not significantly differ between those who lack education and those who are sufficiently educated We argue that education is not a strong determinant of insurance buying, contrary to what such recommendations implicitly assume Simply put, if education does not significantly improve insurance buying, then the benefits from monetary resources and non-monetary efforts devoted to consumer education may be inconsequential We recognize that in some cases, research has demonstrated a negative relationship between health insurance literacy and the likelihood of delayed or foregone care owing to cost for both preventive and non-preventive care.4 However, in most cases, not only is financial illiteracy the norm, but those who are financially literate not show significantly different insurance buying behaviour.5 We offer the insight that individuals who have been in a previous situation of losses (where they could easily fathom the benefits from being insured) and who borrowed from usurious sources to meet medical expenses are significantly more likely to purchase insurance regardless of whether or not they are educated In rural and semi-urban areas, moneylenders and pawnbrokers (who grant credit at exorbitant interest rates) play an essential role during a health crisis The repayment obligation of high-cost credit would also influence enrolment decisions; thus, the households borrowing from usurious sources are more likely to enrol in micro health insurance (MHI) The Indian public health system has not yet caught up with the demand of the population of over a billion because of financial (weak tax compliance, and ineffective tax collection machinery) and human resource constraints Health insurance in India is also under-developed – it is characterized by low levels of government healthcare expenditure (1.18% of gross domestic product) and high out-of-pocket expense (OOPE) that approximately amount to 60.6% of total health expenditure.6 The households in the informal sector fall below the poverty line during illness due to wage loss, catastrophic medical expenses, and repeated medical treatment.7 Thus, iatrogenic poverty (defined as medical illness-induced poverty) often leads to further impoverishment of the already poor households when they resort to financing out of savings, borrowing from informal sources, sale of productive assets, paying from current budget by reducing consumption, substituting or increasing labour supply, or reallocation of resources within the household.8,9 One-fourth of hospitalized Indians fall below the poverty line after a medical treatment, while more than two-fifths of inpatients borrow or sell assets to meet the treatment cost.10 Among these ex-post strategies, informal exploitative credit from money lenders or pawnbrokers, or (sometimes) even microfinance institutions (MFIs) has negative consequences on current financial health and future economic status of households.11 Therefore, the Ayushman Bharat Yojana (National Health Protection Scheme), an ambitious (and so far, the largest) social health insurance programme in the world, was launched in 2018 to provide a coverage of INR 0.5 million (1 USD = approximately INR 71 as on October 2019) for over 10 crore poor families Before this scheme, several non-government organizations or MFIs offered MHI as an extension of existing microcredit activities However, few studies question the financial viability of the schemes owing to a small risk pool, problem of information asymmetry, and excessive reliance on subsidies or external grants.12-16 Poor penetration has been identified as one of the prominent reasons for the failure of MHI, a matter of great concern for low- and middle-income countries.14,17-19 Low uptake of microinsurance has been observed in African countries.16,20 Hence the success of these schemes in achieving universal coverage is debatable if it fails to create value for the poor ensuing lower membership base and limited riskpooling.14,18,21,22 We chose Sampoorna Suraksha Programme (SSP), one of the MHI programmes with largest risk pool in India nested in a broader socio-economic development programme in Karnataka, to understand the determinants of enrolment Literature Review Enrolment is influenced by hospitalizations (often a proxy for health status), perceived self-health status and chronic illness in the household.23 Another study demonstrated that the experience of chronic illnesses in households, education, age, and gender of the head of the households are associated with variation in enrolment.24,25 The households having high ratio of ill members and those reporting chronic illness enrol in MHI.25-27 Most of these studies however, look at how insurance buying behaviour is associated with such socio-demographic variables We try to go a step further and attempt to establish causality More specifically, for example, the logit regressions in the literature assume a well-defined direction of causality from health condition to insurance In general, it must be emphasized that access to insurance may also lead to better well-being in the long run Such mutual feedback effects between (2 or more) variables of interest should be accounted for in any refined statistical analysis Therefore, we use a robust three-stage least squares (3SLS) technique (details explained later), instead of unidirectional logit models to bring in a channel of causality to the existing research In a sense, therefore, our contribution can also be seen as methodological Thus, households that are exposed to higher risk of illness requiring hospitalization or those with higher health expenditure can be expected to enrol in MHI (through that very channel of causality) There is a direct benefit of understanding causality over association: the interplay of so many variables could make the direction of any association look non-specific – a problem that causality directly addresses Indeed, prior research findings that have aimed at understanding how enrolment is associated with other variables, have arrived at diverse (and mixed) conclusions, to which we turn now The households having ill members demand health insurance,27 pay more to participate in insurance scheme,28 and are more likely to renew the policy.29 Individuals with worse health status enrol more than those with better health International Journal of Health Policy and Management, 2021, 10(4), 192–200 193 Savitha and Banerjee status.30 Health expenditure imposes a burden on the income of the household, and thus may positively influence enrolment.31-33 A substantial uptake in health insurance because of escalated healthcare costs has been documented.34 Another study highlights the role of current and future health expenditure, and the perception of future healthcare risks, in health insurance purchase decisions.35 Clearly, the demand for insurance should include an absolute reduction of hardship financing,36 and enrolment is greatly influenced by the desire to reduce this risk of hardship financing Dror and Firth37 argue that individuals incur very high expenses The deficit between insurance cover and medical expenses is often financed by usurious credit So far, adverse selection and its impact on healthcare financing and sustainability have been the focus of earlier studies Our study goes one step further in demonstrating that in addition to illness, associated usurious borrowing determines enrolment Now we explicitly discuss the mixed results on the relation between education and enrolment, that have been highlighted in the literature Income, which could ease some of the hardship financing discussed above, is often directly linked to education Clearly, an educated person can be expected to have a higher income and report a positive association between these variables and enrolment.29,38 This positive association between income (a proxy for affordability) and health insurance purchase is documented by many studies conducted in different countries.39 On the other hand, few studies failed to observe the influence of income in shaping enrolment decisions.40,41 Many studies document a positive association between education and risk aversion and hence, higher demand for insurance.25,42 Individuals with higher levels of education engage in preventive behaviour and appreciate the benefits of insurance as a protective tool, and hence there is a direct relation between education and demand for health insurance.43,44 However, it has been established that a negative association between education and enrolment exists – less educated heads of households are more likely to enrol compared to highly educated heads.27 The logic is that less educated agents, on an average, engage in worse healthcare practices (in comparison to those who are educated) and therefore feel a greater need to remain insured In a nutshell, therefore, the assumptions of established theories on demand for insurance explaining the role of attitude to risk (Friedman and Savage vs Kahneman and Tversky), expected utility (von-Neumann and Morgenstern) and moral hazard45 may not directly hold in large informal economies such as India The validity of many axioms could be undermined in the presence of group consensus and collective good,46 informal mutual insurance,47 low awareness and misinterpretation of information,48 difficulty in enforcing contracts,49 preference for high-frequency events involving uncertain cost over predictable and low cost events and high variance of OOPEs.41,50,51 Refuting the relevance of conventional demand theories for the violation of the underlying assumptions in the informal sector,37 calls for a new approach that states that social capital (group affiliation and 194 reciprocity), imperfect market conditions and the perception that health insurance improves community welfare determine enrollment In the informal sector, gaining access to unaffordable healthcare services during illness is highly valuable, and thus, the health insurance preference of individuals is greatly influenced by current health, past health behaviour, and health investments The enrolment models developed by Ito and Kono25 and Bonan et al52 use household and individual characteristics as a proxy for subjective apprehension and risk behaviour Outreville53 groups the factors determining demand for life insurance under economic, demographic, socio-cultural and institutional categories Akin to a study by Mahmood et al,26 we adopted this framework by incorporating economic factors (income, types of borrowing for medical needs), social factors (education) and demographic factors (illness experience as a proxy for health status), but excluding structural factors such as non-government organization membership given that self-help group (SHG) membership is a prerequisite for buying MHI policy Methods Study Context The SSP was started in 2004 by SKDRDP (Sri Kshetra Dharmasthala Rural Development Project) to provide financial assistance to meet the unexpected medical expenses to the stakeholders and their family, to facilitate access to the best hospitals and to provide medical facilities at lower costs This voluntary membership-based bundled scheme is offered to SHGs and their family members in the age bracket from months to 80 years Enrolment of members takes place in February of every year Sampoorna Suraksha provides medical benefits (health treatment) and exclusive benefits (delivery allowances, death consolation, and domiciliary treatment) The sum assured per member per year is INR 10 000 The scheme offers a family floater cover for members up to INR 70 000, depending on the medical condition and hospital bills The insured members could get the medical treatment in any of the 110-network hospitals with or without referral from another doctor In 2010-2011, 660 185 members from 420 302 families joined SSP, and INR 364 085 225 was mobilized as premium in 2011-2012 A total of INR 45 5493 625 was given as claim benefits to 133 962 individuals in 2010-2011 Study Design This cross-sectional descriptive study was designed to collect quantitative data using survey methodology in the first half of the year 2011 We are primarily interested in the factors that motivate households to join MHI We remain open to the possibility that the demand (for insurance) determinants of newly joined households and those of existing insured households need not be the same The factors that determine enrolment are past illness experience and financial consequences of illness such as borrowing We also controlled for the monthly income of the family; marital status, age, education and occupation of the head of the households; area of residence for descriptive analysis Of these, control factors International Journal of Health Policy and Management, 2021, 10(4), 192–200 Savitha and Banerjee were noteworthy; education (understanding of insurance) and income (affordability of the premium) of the households Borrowing from high-cost (usurious such as money lenders, pawnbrokers) and low-cost sources (non-usurious such as friends, relatives, neighbours) were related to education level of the head of the households If the head is more educated, she can be expected to avoid usurious credit We present our findings using 3SLS regression.54 We looked into loans from money lenders/pawnbrokers/MFIs as informal usurious sources of credit Although MFI is a formal source of finance in India, the credit should be used for productive purposes rather than for consumption smoothing Since the use of MFI credit for health expenses does not generate income that can be used to pay back the debt, we considered MFI credit for medical care as a source that jeopardizes future household consumption with negative consequences Thus, it was clubbed with loan from money lenders and pawnbrokers We coded “0” in the model if credit was taken from non-usurious sources such as neighbours, friends, community members, or relatives The data on SSP membership, illness episodes and subsequent costs of treatment in the previous year of the study, types of borrowing during health shocks, cost of treatment and socio-economic characteristics (age, gender, occupation, education, monthly income, marital status, and area of residence) was collected The questionnaire was piloted and checked for content validity and reliability by using testretest method As the target population size was 892 740 households in 2011-2012 (SSP households were 420 302 that included insured and newly insured), 385 was considered as desirable sample size according to the method of binding frontiers.55 A multi-stage cluster design with random selection procedures was adopted to select households for the study In the first stage, districts where SSP was being implemented were selected, and later, 10 taluks (administrative regions) from these districts were selected based on literacy index In the third stage, 18 valayas (divisions in each taluk) were chosen from these taluks, and later, 84 karyakshetras (villages) were randomly selected from the list given by the project office In the next stage, using the list of households in each karyakshetra, 782 households were selected using systematic sampling method In the sample, 416 were renewed insured, and 366 were newly insured Results Socio-Economic Profile of Households Predominantly, men were found to be heading the households in both groups (newly insured 84.7%; renewed 83.4%) (P = 624) The mean age of household head in newly insured households was 48 (SD 10) years, and that of renewed was 47 years (SD 11) (P = 150) The mean distance to hospitals for renewed households was 2.3 km and for newly insured 2.8 km (P < 05) Each type of household had members on an average (P > 05) The monthly income of renewed insured was INR 8773 (SD INR 7076), and newly insured was INR 9738 (SD INR 9609) (P = 150) The occupation of most heads of the households in renewed insured and newly insured group was daily labour (Table 1) Table Basic Characteristics of Households Renewed Insured (n = 416) Newly Insured (n = 366) Education of head of the household (%) Test Value 4.22 Illiterate 23.1 26.5 Primary 43.5 42.1 Secondary and above 33.4 31.4 Occupation of head of the household (%) 10.01 Waged labourer 56.5 60.4 Home maker/unable to work 6.5 4.9 Self-employment 9.9 5.5 Formal sector employment 3.1 5.7 Unemployed 12.3 12.8 Salaried (informal sector) 8.4 6.8 Agriculture 3.4 3.8 Area of residence 14.88* Rural 7.2 14.2 Urban 40.6 30.6 Semi Urban 52.2 55.2 * P